This repository contains the work done for a SUPINFO school project where students were asked to code a cipher system, the Jefferson Disk, in Python and Pygame for the graphical user interface.
Being a group project, it involved two persons:

• Pierre-Louis DOURNEAU (@pldiiw)
• Clément GRIMAUD (@ehrakis)

Requires Python 3.5+ and Pygame.

Clone the repository:

git clone https://github.com/pldiiw/1ari-mp.git

Then run the GUI:

python3 src/JeffersonGUI.py

There's a test suite inside the test/ directory.

To run it, just execute the test script:

./run_tests.sh

It requires mypy. If don't have it, just run:

pip3 install mypy-lang

All source files can be found in the src/ directory.
The GUI has been splitted into components, each in its own file. All components can be found under the src/component/ directory.
Tests are under the test/ directory.

Coding conventions are important to keep the code base clean. We followed the guidelines defined in the PEP-8. YAPF helped us reach this goal without headaches, and by saving a lot of time.

Function annotations is a new language feature introduced in Python 3.5. It can be used to explicitly type function parameters and variables. Along with Mypy, it helped us avoid bugs ahead of runtime, sketch the flow of the code and easily understand and share function type signatures.

Sharing code and combining with the work of others can be an intricate process. Git hugely facilitates this by providing the appropriate tools to get a more organized workflow and keep track of our advancement.

YAPF is a code formatter for Python that goes further than others linting tools. It totally rearrange the code to get a consistent code style, following a given style guide, like the PEP-8.

Isort is a little utility that alphabetically sort imports and seperate them into sections, making imports more readable and pleasent. Alongside YAPF, it pushes even further the concept of a consistent code style.

The functions and procedures names asked in the subject differ a bit from the ones we wrote. We chose to slightly modify the naming to have a codebase matching the PEP-8 closer and having more expressive names, resulting in more readable code.

Here's a corresponding table with on the left the subroutine name in the subject and on the right the matching name inside our code:

The following sections contains the answers to the questions which are in the subject.

The unencrypted message was : "The quick brown fox jump over the lazy dog".

If we encrypt the text "Hello world" with the cylinder contained in the file cylinder-question.txt and the key [8, 4, 6, 7, 2, 1, 10, 5, 3, 9] we end up with this encrypted message : 'XNUEDNRCHN'.

When exchanging a message using this algorithm, you have also the key and the disks to give. Communicating a key is common to many encryption algorithms, but having another factor to carry, the cylinder, is one more failure spot.

The first weak point of any encryption system using a unique key is exchanging that key: if you encrypt your message, it is therefore that you're afraid of being read by someone else than your peer ; then how can you accept to exchange a key if your communication channel is insecure?

Having the cylinder to also exchange is one more handshake to give. It can seen as more secure, as if your key leaks the theft cannot decrypt the message in any way. But if your cylinder leaks, it is another scenario. The third party has a lot more chance of decrypting your message. He has just to try all the possible permutations of disks, or n! for n disks. For short messages, is it easy to test all the possible keys. The longer the message the harder it is.

Before computers, messages above 6 characters were starting to be very hard to decrypt without the key. 6 characters long messages are already hard to decipher by hand, there's 720 possible keys. Now that we have machines with heavy computational power we can decrypt way longer messages without having the key in a reasonable time.

This algorithm requires no mathematic knowledge for the sender and the recipient to use it, this must have been quite convenient in the 1800s where the average mathematic skills were way lower than the ones someone of our century might have.

What's less convenient about this ciphering system is that is it quite cumbersome to setup and use. You have to exchange the message, the key and the cylinder in order to be able to decrypt the former. That's too much information about our message.

This is a permutation without repetition, therefore there is n! possible keys for n given disks.

This repository is under the Unlicense. See LICENSE file for more information.
EXCEPTING the file subject.pdf which is licensed to SUPINFO International University under the terms of the FreeBSD License.